Infectivity of patients with pulmonary tuberculosis during chemotherapy.

In the current number of the Journal, Ci.ANCY et al. [1] have examined 29 specimens of sputum from patients before and during chemotherapy by direct smear, culture and guinea-pig inoculation with the aim of deciding whether there is any change in pathogenicity during the early weeks of chemotherapy. They found that the occurrence of tuberculosis in the guinea-pig was more closely associated with the results of culture than with the duration of the therapy received by the patient at the time of collection of the specimen and conclude that a change in pathogenicity of the tubercle bacilli as a result of treatment was unlikely to have occurred. They raise the question of whether patients with "sputum positive tuberculosis", despite being on effective chemotherapy, can be regarded as non-infectious, especially as they may be in contact with other immunocompromised patients. Now, "sputum positive" implies a smear-positive or a culturepositive result. There is, however, a high incidence (20~0%) of smear-positive, culture-negative specimens in the first few months of short course therapy [2], perhaps due to retention of acid-fastness by bacilli killed by the antituberculosis drugs. Thus, it would be more appropriate to consider only the infectiousness of culture-positive tuberculosis. The distinction is important, particularly in developing countries, where routine cultures for tubercle bacilli are rarely available. Controlled studies show that sputum culture conversion is very rapid; 85-90% of initially smear-positive patients treated with currently recommended short-course regimens containing rifampicin and pyrazinamide have negative cultures at 2 months and only an occasional scanty positive culture is obtained from any of them thereafter [3~]. What then are the risks of infecting contacts during the initial period of treatment while cultures and guinea-pig inoculation may still be positive? Untreated patients seldom have counts of over 10 colony forming units of Mycobacterium tuberculosis per ml sputum, while 106 is a more usual count for smear-positive patients. At 8 weeks, conversion has generally occurred so that these counts must then have fallen by about 1()6-107 • Data from JINDAI\'1 et al. [7] suggests that sputum counts from patients on regimens containing isoniazid, rifampicin and pyrazinamide fall by about 20-fold in the first 2 days and by a further 200-fold in the next 12 days to reduce the counts of an initially smear-positive patient to about 103 per ml at2 weeks, a level below the estimates of 10ls_10